Update gRPC API with guided decoding and adapter_id changes (#10)

Signed-off-by: Nick Hill <[email protected]>

Author: njhill

proto/generation.proto

@@ -27,15 +27,19 @@ enum DecodingMethod {
 
 message BatchedGenerationRequest {
   string model_id = 1;
+  // Deprecated in favor of adapter_id
   optional string prefix_id = 2;
+  optional string adapter_id = 4;
   repeated GenerationRequest requests = 3;
 
   Parameters params = 10;
 }
 
 message SingleGenerationRequest {
   string model_id = 1;
+  // Deprecated in favor of adapter_id
   optional string prefix_id = 2;
+  optional string adapter_id = 4;
   GenerationRequest request = 3;
 
   Parameters params = 10;
@@ -106,6 +110,31 @@ message DecodingParameters {
   // Exponentially increases the score of the EOS token
   // once start_index tokens have been generated
   optional LengthPenalty length_penalty = 2;
+
+  enum ResponseFormat {
+    // Plain text, no constraints
+    TEXT = 0;
+    // Valid json
+    JSON = 1;
+  }
+
+  message StringChoices {
+    repeated string choices = 1;
+  }
+
+  // Mutually-exclusive guided decoding options
+  oneof guided {
+    // Output will be in the specified format
+    ResponseFormat format = 3;
+    // Output will follow the provided JSON schema
+    string json_schema = 4;
+    // Output will follow the provided regex pattern
+    string regex = 5;
+    // Output will be exactly one of the specified choices
+    StringChoices choice = 6;
+    // Output will follow the provided context free grammar
+    string grammar = 7;
+  }
 }